Polygonum multiflorum disease and insect pest detection method based on YOLOv8

Abstract

The invention discloses a YOLOv8-based polygonum multiflorum disease and insect pest detection method, which comprises the following steps: acquiring a disease and insect pest image of polygonum multiflorum, constructing a disease and insect pest image data set, and marking; inputting the marked polygonum multiflorum disease and insect pest image data set into a YOLOv8-based detection model for training to obtain a polygonum multiflorum disease and insect pest detection model; acquiring real-time polygonum multiflorum leaf and rhizome part images, inputting the polygonum multiflorum pest and disease damage detection model, and marking areas with pest and disease damage; carrying out severity grade division on the disease and pest severity degree of the part; acquiring a polygonum multiflorum lesion process based on the disease and pest images of each stage, and constructing a disease and pest development prediction model; and performing protection measures on the polygonum multiflorum in each state based on the prediction model. The method has the advantages that through the YOLOv8 model and the disease and pest development prediction model, accurate recognition and early warning of diseases and pests of polygonum multiflorum are achieved, a scientific prevention and control basis is provided in combination with disease severity evaluation, and agricultural production efficiency and environmental sustainability are improved.

Description

Technical Field

[0001] The present invention relates to YOLOv8, and in particular to a method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8. Background Art

[0002] Polygonum multiflorum is a commonly used medicinal ingredient in traditional Chinese medicine (TCM), known for its benefits of nourishing the liver and kidneys, slowing aging, and promoting blood circulation. Due to its wide range of pharmacological effects, it has been widely used clinically, particularly in health care and the treatment of some chronic diseases. However, Polygonum multiflorum grows in a complex environment, making it susceptible to various pests and diseases, particularly root rot, leaf lesions, and various insect pests such as aphids and mites. These pests and diseases not only directly affect the growth and development of the plant but can also lead to a reduction in active ingredients and even the production of harmful substances, seriously impacting the quality and safety of the medicinal material. Therefore, efficient and accurate detection and control of pests and diseases are crucial to ensuring the quality of Polygonum multiflorum.

[0003] Currently, the main methods for detecting pests and diseases in Polygonum multiflorum rely on manual inspections, traditional image processing techniques, and some basic machine learning algorithms. However, these methods generally have limitations. Manual inspections rely on manual identification, which is time-consuming and labor-intensive, and is susceptible to subjective factors. This can easily overlook early signs of pests and diseases, resulting in a delay in identifying problems. Traditional image processing methods often rely on manual feature extraction, which is ineffective for complex pest and disease manifestations and is sensitive to image noise and background interference, resulting in low accuracy. Although simple machine learning methods have been used for pest and disease identification in some static images, they lack the efficiency to process real-time monitoring data and are unable to adapt to the dynamic changes of pests and diseases at different growth stages. Furthermore, most traditional methods can only identify a single type of pest and disease and have limited comprehensive detection and location capabilities for multiple pests and diseases, making them difficult to simultaneously address multiple pests and diseases, such as leaf spot, root rot, aphids, and spider mites. More importantly, traditional methods lack the ability to predict pest and disease development trends, making it difficult to effectively grasp the potential threats posed by pests and diseases, and unable to carry out scientific and precise prevention and control. Summary of the Invention

[0004] To improve existing methods for detecting diseases and insect pests in Polygonum multiflorum, this paper presents a method based on YOLOv8. This method uses the YOLOv8 model to achieve real-time detection and accurate identification of Polygonum multiflorum diseases and insect pests, and combines this with disease severity assessment to provide a scientific basis for prevention and control. By constructing a model to predict the development of pests and diseases, early warnings can be provided and precise protective measures can be implemented, improving agricultural production efficiency and environmental sustainability.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8, comprising:

[0007] Obtain images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, and construct a dataset of Polygonum multiflorum pest and disease images. Label images of four types of targets: leaf spot, root rot, aphids, and spider mites.

[0008] Based on the labeled Polygonum multiflorum disease and insect pest image dataset, it is input into the YOLOv8-based detection model for training, and the output is the trained Polygonum multiflorum disease and insect pest detection model;

[0009] Acquire real-time images of Polygonum multiflorum leaves and rhizomes, identify and detect the images based on the trained Polygonum multiflorum pest and disease detection model, and mark areas with pests and diseases;

[0010] Based on the types of pests and diseases, area and depth parameters of the marked pest and disease areas, the severity of the pests and diseases in the area is divided into severity levels;

[0011] Based on images of pests and diseases at various stages, the pathological process of Polygonum multiflorum is obtained and a pest and disease development prediction model is constructed;

[0012] Based on the pest and disease development prediction model, protective measures are implemented for Polygonum multiflorum that is about to develop disease or has already developed disease.

[0013] Preferably, the acquisition of pest and disease images of the leaves and rhizomes of Polygonum multiflorum, the construction of a Polygonum multiflorum pest and disease image dataset, and the labeling of images containing four types of targets, namely leaf spot, root rot, aphids, and red spider mites, specifically include:

[0014] Based on the acquired images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, a dataset of Polygonum multiflorum pests and diseases was constructed.

[0015] Perform image screening and format unification on the images in the Polygonum multiflorum disease and insect pest image dataset;

[0016] Based on the processed data and images, the diseased parts of the leaves and rhizomes of Polygonum multiflorum were annotated with polygons using LabelMe.

[0017] Preferably, the labeled Polygonum multiflorum disease and insect pest image dataset is input into a detection model based on YOLOv8 for training, and the output is a trained Polygonum multiflorum disease and insect pest detection model, which specifically includes:

[0018] Based on the labeled Polygonum multiflorum disease and insect pest image dataset, more image data for training is obtained by geometric transformation and color perturbation of the images.

[0019] The original image data set and enhanced image data are input into the detection model based on YOLOv8 for training, and a Polygonum multiflorum pest and disease detection model is obtained;

[0020] Based on the trained Polygonum multiflorum pest and disease detection model, the verification image set is input into the model to obtain the detection results;

[0021] Based on the detection model results and the pest and disease type and pest and disease area depth parameter data in the validation image set, if they are the same, the trained Polygonum multiflorum pest and disease detection model is output;

[0022] If they are not the same, the pest and disease type and pest and disease area depth parameters in the verification image set are marked, and the model is retrained until the training detection model results are the same as the pest and disease type and pest and disease area depth parameter data in the verification image set.

[0023] Preferably, the real-time images of the leaves and rhizomes of Polygonum multiflorum are acquired, and based on the trained Polygonum multiflorum pest and disease detection model, the images are identified and detected, and the areas with pests and diseases are marked, specifically including:

[0024] Multispectral imaging equipment is used to collect images of Polygonum multiflorum leaves and rhizomes in real time, and the images are screened and formatted in a unified manner.

[0025] Based on the processed Polygonum multiflorum image, the image is input into a Polygonum multiflorum disease and insect pest detection model to obtain a Polygonum multiflorum image with marked disease and insect pest parts;

[0026] Based on the labeled images of Polygonum multiflorum with pest and disease parts, the pest and disease types and the area and depth parameter data of the pest and disease areas of each image are obtained.

[0027] Preferably, the method of classifying the severity of the pests in the marked area based on the pest type, area and depth parameters specifically includes:

[0028] The weights of pests and diseases were set based on the degree of their impact on the growth of Polygonum multiflorum, and the area depth data were standardized;

[0029] A severity scoring model was constructed based on the pest and disease type weights and the standardized area-depth data;

[0030] Based on the scoring results given by the severity scoring model, the real-time acquired Polygonum multiflorum diseased edge images were divided into severity levels.

[0031] Preferably, the acquisition of the pathological process of Polygonum multiflorum based on the pest and disease images at each stage and the construction of the pest and disease development prediction model specifically include:

[0032] Based on the pest and disease images of the same plant at different stages under different pest and disease types, the area spread rate, depth deterioration rate, and pest and disease density growth rate data are obtained, and time series data is formed;

[0033] Based on time series data, feature extraction is performed on image features and time series features to build a pest and disease development prediction model.

[0034] Preferably, the protective measures implemented on Polygonum multiflorum that are about to develop lesions or have already developed lesions based on the pest and disease development prediction model specifically include:

[0035] Based on the acquired pest and disease development prediction model, real-time images of Polygonum multiflorum leaves and rhizomes are input into it to determine whether pests and diseases have occurred and to predict future pathological conditions;

[0036] Based on the prediction results, biological control intervention and environmental regulation are carried out in the diseased areas.

[0037] Compared with the prior art, the advantages of the present invention are:

[0038] By constructing and annotating a specialized dataset of pest and disease images, we can provide rich and targeted image data for training models, ensuring the high accuracy of detection models in identifying pests and diseases.

[0039] As an efficient target detection algorithm, YOLOv8 can quickly and accurately identify pest and disease areas in real-time images, greatly improving the efficiency of pest and disease monitoring and reducing the workload of manual inspections.

[0040] Secondly, by marking pest and disease areas in detail and combining them with the classification of disease severity, a comprehensive assessment of the development stage of pests and diseases can be achieved, thus providing a scientific basis for precise prevention and control.

[0041] In addition, the construction of a pest and disease development prediction model makes the prevention of pests and diseases more accurate, and can detect potential risks in advance and take protective measures to reduce the harm of pests and diseases from the source.

[0042] This intelligent pest and disease monitoring and control method can improve the yield and quality of crops, reduce the use of chemical pesticides, and have a positive impact on the ecological environment. It is an innovative solution for pest and disease management in modern agriculture. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of the method proposed in the present invention;

[0044] Figure 2 This is a schematic diagram of image annotation proposed by the present invention;

[0045] Figure 3This is a schematic diagram of the construction of the Polygonum multiflorum disease and insect pest detection model proposed in the present invention;

[0046] Figure 4 This is a schematic diagram of the lesion area identification, detection and annotation proposed by the present invention;

[0047] Figure 5 This is a schematic diagram of the severity level classification proposed by the present invention;

[0048] Figure 6 This is a schematic diagram of the pest and disease development prediction model proposed in the present invention;

[0049] Figure 7 A schematic diagram of the protection measures proposed by the present invention;

[0050] Figure 8 This is a diagram of the architecture of the electronic equipment in this solution;

[0051] Figure 9 This is a schematic diagram of the computer-readable storage medium structure in this solution. DETAILED DESCRIPTION

[0052] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0053] See Figure 1 As shown, a method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 includes:

[0054] Step 1: Obtain images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, and construct a dataset of Polygonum multiflorum pest and disease images. Label images of four types of targets: leaf spot, root rot, aphids, and spider mites.

[0055] Step 2: Based on the labeled Polygonum multiflorum disease and insect pest image dataset, input it into the YOLOv8-based detection model for training, and output the trained Polygonum multiflorum disease and insect pest detection model;

[0056] Step 3: Obtain real-time images of Polygonum multiflorum leaves and rhizomes. Based on the trained Polygonum multiflorum pest and disease detection model, identify and detect the images and mark areas with pests and diseases.

[0057] Step 4: Based on the pest and disease type, area and depth parameters of the marked pest and disease area, the severity of the pest and disease in the area is classified into severity levels;

[0058] Step 5: Based on the images of pests and diseases at each stage, the pathological process of Polygonum multiflorum is obtained and a pest and disease development prediction model is constructed;

[0059] Step 6: Implement protective measures for Polygonum multiflorum that is about to develop disease or has already developed disease based on the pest and disease development prediction model.

[0060] See Figure 2 As shown in the figure, we obtain images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, and construct a dataset of Polygonum multiflorum pests and diseases images. We label images of four types of targets, including leaf spot, root rot, aphids, and spider mites. The specific steps include:

[0061] Based on the acquired images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, a dataset of Polygonum multiflorum pests and diseases was constructed.

[0062] Perform image screening and format unification on the images in the Polygonum multiflorum disease and insect pest image dataset;

[0063] Based on the processed data and images, the diseased parts of the leaves and rhizomes of Polygonum multiflorum were annotated with polygons using LabelMe.

[0064] Specifically, the labeling rules for various diseases and pests of Polygonum multiflorum include:

[0065] Leaf spot: Mark irregular spots on leaves (avoiding veins). If multiple spots are adjacent and the distance between them is less than 5mm, merge them and mark them together, and keep the healthy vein area.

[0066] Root rot: Mark the rotten area of the root (brown / black part). If the root is completely rotten, mark the entire root (the plant needs to be dug up and the roots need to be photographed).

[0067] Aphids mark the minimum enclosing rectangle (covering antennae) of a single insect or dense insect swarm.

[0068] Spider mites: Mark the egg clusters (red dotted areas) and adult spider mites (oval bodies) on the underside of the leaves.

[0069] If the lesion areas overlap, they are divided into multiple polygons for annotation. Based on the annotated data set, the training set, validation set, and test set are divided in a ratio of 7:2:1 to ensure balanced category distribution.

[0070] See Figure 3 As shown in the figure, based on the labeled Polygonum multiflorum disease and insect pest image dataset, the YOLOv8-based detection model is input for training, and the trained Polygonum multiflorum disease and insect pest detection model is output. Specifically, it includes:

[0071] Based on the labeled Polygonum multiflorum disease and insect pest image dataset, more image data for training is obtained by geometric transformation and color perturbation of the images.

[0072] The original image data set and enhanced image data are input into the detection model based on YOLOv8 for training, and a Polygonum multiflorum pest and disease detection model is obtained;

[0073] Based on the trained Polygonum multiflorum pest and disease detection model, the verification image set is input into the model to obtain the detection results;

[0074] Based on the detection model results and the pest and disease type and pest and disease area depth parameter data in the validation image set, if they are the same, the trained Polygonum multiflorum pest and disease detection model is output;

[0075] If they are not the same, the pest and disease type and pest and disease area depth parameters in the verification image set are marked, and the model is retrained until the training detection model results are the same as the pest and disease type and pest and disease area depth parameter data in the verification image set.

[0076] Specifically, when performing geometric transformation on the image of Polygonum multiflorum pests and diseases, for the selected bounding box vertices (x i ,y i ), the coordinate correction formula is:

[0077] x′ i =x c +(x i -x c )cosθ-(y i -y c )sinθ

[0078] y′ i =y c +(x i -x c )sinθ+(y i -y c )cosθ

[0079] Where θ is the rotation angle, (x c ,y c ) is the center point of the image.

[0080] The main goal of model training is to minimize the loss function, which includes position loss, confidence loss and category loss. Specifically,

[0081] The position loss is used to calculate the difference between the predicted box and the true box, and the formula is:

[0082]

[0083] in, is the center point and width and height of the prediction box, x i 、y i 、w i 、h i is the corresponding value of the real box;

[0084] Confidence loss is used to measure the difference between the model's confidence in the presence of pests and diseases and the actual situation. The formula is:

[0085]

[0086] in, is the target confidence of the model prediction, C i is the confidence of the true label;

[0087] Class loss is used to measure the difference between the pest and disease type predicted by the model and the actual type. The formula is:

[0088]

[0089] in, is the probability of pest and disease category predicted by the model, P i is the true category label;

[0090] By minimizing the total loss function L = L loc +L conf +L class To train the YOLOv8 model.

[0091] See Figure 4 As shown, real-time images of the leaves and rhizomes of Polygonum multiflorum are obtained. Based on the trained Polygonum multiflorum pest and disease detection model, the images are identified and detected, and the areas with pests and diseases are marked, including:

[0092] Multispectral imaging equipment is used to collect images of Polygonum multiflorum leaves and rhizomes in real time, and the images are screened and formatted in a unified manner.

[0093] Based on the processed Polygonum multiflorum image, the image is input into a Polygonum multiflorum disease and insect pest detection model to obtain a Polygonum multiflorum image with marked disease and insect pest parts;

[0094] Based on the labeled images of Polygonum multiflorum with pest and disease parts, the pest and disease types and the area and depth parameter data of the pest and disease areas of each image are obtained.

[0095] Specifically, the collected multispectral images may contain some invalid images (such as blurred, noisy or overexposed images). Clear images are screened using the Laplace transform algorithm. For each image I:

[0096]

[0097] in, and is the image gradient, and N is the number of image pixels.

[0098] The inference process of the YOLOv8 model is carried out through a convolutional neural network, which outputs the location box (bounding box), category label and confidence of the pest area in the image. For each detected pest area, YOLOv8 will give the category information to indicate which pest type the area belongs to, the confidence information to indicate the model's confidence in whether the area is indeed a pest, and the bounding box location information to indicate the coordinates of the upper left corner of the target area (x min ,y min ) and the lower right corner coordinate (x max ,y max ), the area information is calculated by the predicted bounding box size, and the area A is: A=(x max -x min )·(y max ,y min ).

[0099] See Figure 5 The severity of the pests and diseases in the marked area is divided into the following levels based on the types of pests and diseases, area and depth parameters:

[0100] The weights of pests and diseases were set based on the degree of their impact on the growth of Polygonum multiflorum, and the area depth data were standardized;

[0101] A severity scoring model was constructed based on the pest and disease type weights and the standardized area-depth data;

[0102] Based on the scoring results given by the severity scoring model, the real-time acquired Polygonum multiflorum diseased edge images were divided into severity levels.

[0103] Specifically, the severity scoring model needs to comprehensively consider the weight of the pest and disease type and the standardized data of area and depth. The comprehensive score S can be calculated using the following formula:

[0104] S=W discase ×(A norm +D norm )

[0105] Among them, W discase is the weight of the pest and disease type, A norm is the standardized area data, D norm For standardized depth data, the weights of pest and disease types can be determined through expert knowledge, literature research or historical data. Specifically, reference can be made to the effects of pests and diseases on the growth of Polygonum multiflorum, such as the effects on growth rate, root development, and yield.

[0106] For the pest and disease conditions that may occur in different locations of the same Polygonum multiflorum plant, the refinement accuracy of the scoring model can be further improved. By dividing the weights of different parts and areas, the scoring formula can be refined as follows:

[0107] S=W discase ×(W leaf ×A leaf +W root ×A rot +D norm )

[0108] Based on the severity score results, the impact of pests and diseases is divided into several levels, specifically:

[0109] Mild (0≤S<30): The impact of pests and diseases is relatively light, and the growth of Polygonum multiflorum is not significantly affected.

[0110] Moderate (30≤S<60): The impact of pests and diseases is moderate, the growth of Polygonum multiflorum is affected to a certain extent, and attention should be paid.

[0111] Severe (60≤S<90): The impact of pests and diseases is serious, the growth of Polygonum multiflorum is significantly affected, and intervention is needed.

[0112] Extremely severe (90≤S≤100): The impact of pests and diseases is serious, the growth of Polygonum multiflorum has almost stopped or has declined extremely, and immediate measures need to be taken.

[0113] See Figure 6 As shown in the figure, based on the images of pests and diseases at each stage, the pathological process of Polygonum multiflorum is obtained and a pest and disease development prediction model is constructed, which specifically includes:

[0114] Based on the pest and disease images of the same plant at different stages under different pest and disease types, the area spread rate, depth deterioration rate, and pest and disease density growth rate data are obtained, and time series data is formed;

[0115] Based on time series data, feature extraction is performed on image features and time series features to build a pest and disease development prediction model.

[0116] Specifically, pest and disease images of the same plant are collected at different time points. Each time the images are collected, the timestamp is calibrated to form time series data. For each image, the characteristic parameters of area spread A, depth deterioration D, and pest and disease density M are extracted, and the growth rate R of each item is calculated. A (t), R D (t), R M (t), the calculated area diffusion rate, depth deterioration rate and pest density growth rate are combined into time series data: {R A (t), R D (t), R M(t)}, t=t1,t2,...,t n .

[0117] The time series data of pests and diseases were modeled and predicted by using LSTM network, and the accuracy of the model was evaluated by cross-validation, mean square error (MSE), mean absolute error (MAE), etc.

[0118] See Figure 7 As shown in the figure, based on the pest and disease development prediction model, the protective measures for Polygonum multiflorum that is about to develop disease or has already developed disease include:

[0119] Based on the acquired pest and disease development prediction model, real-time images of Polygonum multiflorum leaves and rhizomes are input into it to determine whether pests and diseases have occurred and to predict future pathological conditions;

[0120] Based on the prediction results, biological control intervention and environmental regulation are carried out in the diseased areas.

[0121] Specifically, biological control intervention involves using image segmentation to precisely locate pest and disease areas, including leaves and rhizomes, to determine the distribution of the disease and pest. Based on the predicted severity of the pest, appropriate biological control measures are selected. Common biological control measures include using natural enemies (such as predatory insects and parasitic wasps) to suppress pests and spraying biological pesticides (such as silicones, microbial pesticides, and natural plant extracts).

[0122] Environmental regulation primarily involves adjusting environmental factors such as temperature, humidity, and light to mitigate the development of pests and diseases. Specific measures include controlling greenhouse temperature, maintaining appropriate humidity, reducing the occurrence of pests and diseases, reducing excessive light intensity, optimizing the plant growth environment, and mitigating the spread of pests and diseases. Precise environmental regulation is also implemented by combining factors such as climatic conditions and crop density.

[0123] Furthermore, the method according to the embodiment of the present application can also be used with the aid of Figure 8 The electronic device architecture shown in FIG. Figure 8 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store a method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 provided in this application. The electronic device 500 may also include a terminal interface 508. Of course, Figure 8 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 8 One or more components of an electronic device are shown.

[0124] Figure 9 This is a schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Figure 9 As shown, a computer-readable storage medium 600 according to one embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, a method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0125] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0126] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8, characterized in that: include: Obtain images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, and construct a dataset of Polygonum multiflorum pest and disease images. Label images of four types of targets: leaf spot, root rot, aphids, and spider mites. Based on the labeled Polygonum multiflorum disease and insect pest image dataset, it is input into the YOLOv8-based detection model for training, and the output is the trained Polygonum multiflorum disease and insect pest detection model; Acquire real-time images of Polygonum multiflorum leaves and rhizomes, identify and detect the images based on the trained Polygonum multiflorum pest and disease detection model, and mark areas with pests and diseases; Based on the types of pests and diseases, area and depth parameters of the marked pest and disease areas, the severity of the pests and diseases in the area is divided into severity levels; Based on images of pests and diseases at various stages, the pathological process of Polygonum multiflorum is obtained and a pest and disease development prediction model is constructed; Based on the pest and disease development prediction model, protective measures are implemented for Polygonum multiflorum that is about to develop disease or has already developed disease.

2. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to claim 1, characterized in that, The acquisition of pest and disease images of the leaves and rhizomes of Polygonum multiflorum, and the construction of a Polygonum multiflorum pest and disease image dataset, and the labeling of images containing four types of targets, namely leaf spot disease, root rot disease, aphids, and red spider mites, specifically include: Based on the acquired images of pests and diseases on the leaves and rhizomes of Polygonum multiflorum, a dataset of Polygonum multiflorum pests and diseases was constructed. Perform image screening and format unification on the images in the Polygonum multiflorum disease and insect pest image dataset; Based on the processed data and images, the diseased parts of the leaves and rhizomes of Polygonum multiflorum were annotated with polygons using LabelMe.

3. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to claim 1, characterized in that, The labeled Polygonum multiflorum disease and insect pest image dataset is input into a detection model based on YOLOv8 for training, and the output is a trained Polygonum multiflorum disease and insect pest detection model, which specifically includes: Based on the labeled Polygonum multiflorum disease and insect pest image dataset, more image data for training is obtained by geometric transformation and color perturbation of the images. The original image data set and enhanced image data are input into the detection model based on YOLOv8 for training, and a Polygonum multiflorum pest and disease detection model is obtained; Based on the trained Polygonum multiflorum pest and disease detection model, the verification image set is input into the model to obtain the detection results; Based on the detection model results and the pest and disease type and pest and disease area depth parameter data in the validation image set, if they are the same, the trained Polygonum multiflorum pest and disease detection model is output; If they are not the same, the pest and disease type and pest and disease area depth parameters in the verification image set are marked, and the model is retrained until the training detection model results are the same as the pest and disease type and pest and disease area depth parameter data in the verification image set.

4. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to claim 1, characterized in that, The real-time images of the leaves and rhizomes of Polygonum multiflorum are obtained, and based on the trained Polygonum multiflorum pest and disease detection model, the images are identified and detected, and the areas with pests and diseases are marked, specifically including: Multispectral imaging equipment is used to collect images of Polygonum multiflorum leaves and rhizomes in real time, and the images are screened and formatted in a unified manner. Based on the processed Polygonum multiflorum image, the image is input into a Polygonum multiflorum disease and insect pest detection model to obtain a Polygonum multiflorum image with marked disease and insect pest parts; Based on the labeled images of Polygonum multiflorum with pest and disease parts, the pest and disease types and the area and depth parameter data of the pest and disease areas of each image are obtained.

5. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to claim 1, characterized in that, The severity level of the pests in the marked area based on the pest type, area and depth parameters specifically includes: The weights of pests and diseases were set based on the degree of their impact on the growth of Polygonum multiflorum, and the area depth data were standardized; A severity scoring model was constructed based on the pest and disease type weights and the standardized area-depth data; Based on the scoring results given by the severity scoring model, the real-time acquired Polygonum multiflorum diseased edge images were divided into severity levels.

6. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to claim 1, characterized in that, The method of obtaining the pathological process of Polygonum multiflorum based on the images of pests and diseases at each stage and constructing the pest and disease development prediction model specifically includes: Based on the pest and disease images of the same plant at different stages under different pest and disease types, the area spread rate, depth deterioration rate, and pest and disease density growth rate data are obtained, and time series data is formed; Based on time series data, feature extraction is performed on image features and time series features to build a pest and disease development prediction model.

7. A method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 according to claim 1, characterized in that, The protective measures for Polygonum multiflorum that are about to develop lesions or have already developed lesions based on the pest and disease development prediction model specifically include: Based on the acquired pest and disease development prediction model, real-time images of Polygonum multiflorum leaves and rhizomes are input into it to determine whether pests and diseases have occurred and to predict future pathological conditions; Based on the prediction results, biological control intervention and environmental regulation are carried out in the diseased areas.

8. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the Polygonum multiflorum pest and disease detection method based on YOLOv8 as described in any one of claims 1-6.

9. A computer-readable storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by a processor, a method for detecting diseases and insect pests of Polygonum multiflorum based on YOLOv8 is implemented according to any one of claims 1 to 6.

Images